Intracardiac Electrogram Targets for Ventricular Tachycardia Ablation.

The pathogenesis of ventricular tachycardia (VT) in most patients with a prior myocardial scarring is reentry involving compartmentalized muscle fibers protected within the scar. Often the 12-lead ECG morphology of the VT itself is not available when treated with a defibrillator. Consequently, VT ablation takes on an interesting challenge of finding critical targets in sinus rhythm. High-density recordings are essential to evaluate a substrate based on whole electrogram voltage and activation delay, supplemented with substrate perturbation through alternate site pacing or introducing an extra stimulation. In this article, we discuss contemporary intracardiac electrogram targets for VT ablation, with explanation on each of their specific fundamental physiology.

The Role of Artificial Intelligence and Machine Learning in Clinical Cardiac Electrophysiology.

In recent years, numerous applications for artificial intelligence (AI) in cardiology have been found, due in part to large digitized data sets and the evolution of high-performance computing. In the discipline of cardiac electrophysiology (EP), a number of clinical, imaging, and electrical waveform data are considered in the diagnosis, prognostication, and management of arrhythmias, which lend themselves well to automation through AI. But equally relevant, AI offers a unique opportunity to discover novel EP concepts and improve clinical care through its inherent, hierarchical tenets of self-learning. In this review we focus on the application of AI in clinical EP and summarize state-of-the art, large, clinical studies in the following key domains: (1) electrocardiogram-based arrhythmia and disease classification; (2) atrial fibrillation source detection; (3) substrate and risk assessment for atrial fibrillation and ventricular tachyarrhythmias; and (4) predicting outcomes after cardiac resynchronization therapy. Many are small, single-centre, proof-of-concept investigations, but they still show ground-breaking performance of deep learning, a subdomain of AI, which surpasses traditional statistical analysis. Larger studies, for instance classifying arrhythmias from electrocardiogram recordings, have further provided external validation of their high accuracy. Ultimately, the performance of AI is dependent on the quality of the input data and the rigour of algorithm development. The field is still nascent and several barriers will need to be overcome, including prospective validation in large, well labelled data sets and more seamless information technology-based data collection/integration, before AI can be adopted into broader clinical EP practice. This review concludes with a discussion of these challenges and future work.

Unipolar electrogram-based voltage mapping with far-field cancellation to improve detection of abnormal atrial substrate during atrial fibrillation.

INTRODUCTION: An important substrate for atrial fibrillation (AF) is fibrotic atrial myopathy. Identifying low voltage, myopathic regions during AF using traditional bipolar voltage mapping is limited by the directional dependency of wave propagation. Our objective was to evaluate directionally independent unipolar voltage mapping, but with far-field cancellation, to identify low-voltage regions during AF. METHODS: In 12 patients undergoing pulmonary vein isolation for AF, high-resolution voltage mapping was performed in the left atrium during sinus rhythm and AF using a roving 20-pole circular catheter. Bipolar electrograms (EGMs) (Bi) < 0.5 mV in sinus rhythm identified low-voltage regions. During AF, bipolar voltage and unipolar voltage maps were created, the latter with (uni-res) and without (uni-orig) far-field cancellation using a novel, validated least-squares algorithm. RESULTS: Uni-res voltage was ~25% lower than uni-orig for both low voltage and normal atrial regions. Far-field EGM had a dominant frequency (DF) of 4.5-6.0 Hz, and its removal resulted in a lower DF for uni-orig compared with uni-res (5.1 ± 1.5 vs. 4.8 ± 1.5 Hz; p < .001). Compared with Bi, uni-res had a significantly greater area under the receiver operator curve (0.80 vs. 0.77; p < .05), specificity (86% vs. 76%; p < .001), and positive predictive value (43% vs. 30%; p < .001) for detecting low-voltage during AF. Similar improvements in specificity and positive predictive value were evident for uni-res versus uni-orig. CONCLUSION: Far-field EGM can be reliably removed from uni-orig using our novel, least-squares algorithm. Compared with Bi and uni-orig, uni-res is more accurate in detecting low-voltage regions during AF. This approach may improve substrate mapping and ablation during AF, and merits further study.

Focal and pseudo/rotational activations in human atrial fibrillation defined with automated periodicity mapping.

INTRODUCTION: Defining atrial fibrillation (AF) wave propagation is challenging unless local signal features are discrete or periodic. Periodic focal or rotational activity may identify AF drivers. Our objective was to characterize AF propagation at sites with periodic activation to evaluate the prevalence and relationship between focal and rotational activation. METHODS: We included 80 patients (61 ± 10 years, persistent AF 49%) from the FaST randomized trial that compared the efficacy of adjunctive focal site ablation versus pulmonary vein isolation. Patients underwent left atrial (LA) activation mapping with a 20-pole circular catheter during spontaneous or induced AF. Five-second bipolar and unipolar electrograms in AF were analyzed. Periodic sites were identified by spectral analysis of the bipolar electrogram. Activation maps of periodic sites were constructed using an automated, validated tracking algorithm, and classified into three patterns: focal sites (FS), rotation (RO), or pseudo-rotation (pRO). RESULTS: The most common propagation pattern at periodic sites was FS for 5-s in all patients (4.9 ± 1.9 per patient). RO and pRO were observed in two and seven patients, respectively, but were all transient (3-5 cycles). Activation from a FS evolved into transient RO/pRO in five patients. No patient had autonomous RO/pRO activations. Patients with RO/pRO had greater LA surface area with periodicity (78 ± 7 vs. 63 ± 16%, p = .0002) and shorter LA periodicity CL (166 ± 10 vs. 190±28 ms, p = .0001) than the rest. CONCLUSION: Using automated, regional AF periodicity mapping, FS is more prevalent and temporally stable than RO/pRO. Most RO/pRO evolve from neighboring FS. These findings and their implications for AF maintenance require verification with global, panoramic mapping.

Omnipolarity applied to equi-spaced electrode array for ventricular tachycardia substrate mapping.

AIMS: Bipolar electrogram (BiEGM)-based substrate maps are heavily influenced by direction of a wavefront to the mapping bipole. In this study, we evaluate high-resolution, orientation-independent peak-to-peak voltage (Vpp) maps obtained with an equi-spaced electrode array and omnipolar EGMs (OTEGMs), measure its beat-to-beat consistency, and assess its ability to delineate diseased areas within the myocardium compared against traditional BiEGMs on two orientations: along (AL) and across (AC) array splines. METHODS AND RESULTS: The endocardium of the left ventricle of 10 pigs (three healthy and seven infarcted) were each mapped using an Advisor™ HD grid with a research EnSite Precision™ system. Cardiac magnetic resonance images with late gadolinium enhancement were registered with electroanatomical maps and were used for gross scar delineation. Over healthy areas, OTEGM Vpp values are larger than AL bipoles by 27% and AC bipoles by 26%, and over infarcted areas OTEGM Vpp values are 23% larger than AL bipoles and 27% larger than AC bipoles (P < 0.05). Omnipolar EGM voltage maps were 37% denser than BiEGM maps. In addition, OTEGM Vpp values are more consistent than bipolar Vpps showing less beat-by-beat variation than BiEGM by 39% and 47% over both infarcted and healthy areas, respectively (P < 0.01). Omnipolar EGM better delineate infarcted areas than traditional BiEGMs from both orientations. CONCLUSION: An equi-spaced electrode grid when combined with omnipolar methodology yielded the largest detectable bipolar-like voltage and is void of directional influences, providing reliable voltage assessment within infarcted and non-infarcted regions of the heart.

Information theory to tachycardia therapy: electrogram entropy predicts diastolic microstructure of reentrant ventricular tachycardia.

There is no known strategy to differentiate which multicomponent electrograms in sinus rhythm maintain reentrant ventricular tachycardia (VT). Low entropy in the voltage breakdown of a multicomponent electrogram can localize conditions suitable for reentry but has not been validated against the classic VT activation mapping. We examined whether low entropy in a late and diversely activated ventricular scar region characterizes and differentiates the diastolic path of VT and represents protected tissue channels devoid of side branches. Intraoperative bipolar electrogram (BiEGM) activation and entropy maps were obtained during sinus rhythm in 17 patients with ischemic cardiomyopathy and compared with diastolic activation paths of VT (total of 39 VTs). Mathematical modeling of a zigzag main channel with side branches was also used to further validate structural representation of low entropy in the ventricular scar. A median of one region per patient (range: 1-2 regions) was identified in sinus rhythm, in which BiEGM with the latest mean activation time and adjacent minimum entropy were assembled together in a high-activation dispersion region. These regions accurately recognized diastolic paths of 34 VTs, often to multiple inducible VTs within a single individual arrhythmogenic region. In mathematical modeling, side branching from the main channel had a strong influence on the BiEGM composition along the main channel. The BiEGM obtained from a long unbranched channel had the lowest entropy compared with those with multiple side branches. In conclusion, among a population of multicomponent sinus electrograms, those that demonstrate low entropy and are delayed colocalize to critical long-protected channels of VT. This information is pertinent for planning VT ablation in sinus rhythm. NEW & NOTEWORTHY Entropy is a measure to quantify breakdown in information. Electrograms from a protected tissue channel can only possess a few states in their voltage and thus less information. In contrast, current-load interactions from side branches in unprotected channels introduce a number of dissimilar voltage deflections and thus high information. We compare here a mapping approach based on entropy against a rigorous reference standard of activation mapping during VT and entropy was assessed in sinus rhythm.

Differential pacing from two sites to diagnose risk of ventricular arrhythmia and death.

BACKGROUND: QRS abnormalities may not be apparent in sinus rhythm in electrically stable cardiomyopathy patients who can have quiescent but highly arrhythmogenic substrate. Here, we test the hypothesis that differential changes in QRS construction during right-ventricular apex pacing (RVP) as opposed to atrial pacing (AP) will identify latent substrate for ventricular arrhythmias (VA) and death. METHODS: Forty patients with cardiomyopathy free of VA underwent baseline 114-electrode body-surface electrocardiogram during AP (100 beats per minute [bpm]) and RVP (100 and 120 bpm). The filtered-averaged QRS at each electrode was deconstructed into individual intra-QRS and post-QRS ventricular myopotentials (VMP ). The primary outcome was VA or death. Prognostic accuracy of VMP was validated using V1 to V6 leads in another prospective cohort of 44-cardiomyopathy patients. RESULTS: Twenty-six patients were eligible for initial analysis. After 5 ± 2 years of follow-up, eight (31%) patients had VA (VAPos ) while rest were uneventful (VANeg ). During AP100 , VAPos patients expressed more VMP than VANeg patients (16 ± 1 vs 12 ± 1, P = 0.02). RVP100 and RVP120 in VAPos patients introduced an additional 5.5 ± 0.5 and 6.0 ± 0.5 VMP (P < 0.0001 vs AP100 ). The relative change with RVP120 versus AP100 in VANeg patients exceeded VAPos patients by 1.2 ± 0.5 VMP (P = 0.03). Increment in VMP count of <8 in lead-V5 with RVP120 compared to AP100 best predicted VA (area under curve 0.81, P = 0.01). In the validation cohort, primary outcome occurred in 13 (33%) patients. Native QRS features and AP100 alone failed to predict primary outcome. Patients with increment in VMP count of <8 in lead-V5 with RVP120 versus AP100 had 7.9-fold increased risk of primary outcome (95% confidence interval 1.01, 61.61; P = 0.049). CONCLUSION: Cardiomyopathy patients at risk of VA or death perturb the QRS less than low-risk patients with differential pacing. This functional response may be useful to identify arrhythmogenic substrate.

Determinants of atrial bipolar voltage: Inter electrode distance and wavefront angle.

BACKGROUND: Local bipolar electrogram (EGM) peak-to-peak voltage (Vpp) is currently used to characterise mapped myocardial substrate. However, how interelectrode distance and angle of wavefront incidence affect bipolar, Vpp values, in the current era of multi-electrode mapping is unknown. OBJECTIVES: To elucidate the effects of tissue and electrode geometry on bipolar Vpp measurements, when mapping healthy versus diseased atrial regions. METHODS: A bidomain model of human atrial tissue was used to quantify the influence on Vpp values of various electrode configurations in healthy tissue, and tissue containing an unexcitable region. The orientation angle and interelectrode spacing of a surface bipole, and thickness and depth of the unexcitable core were serially varied. Results were validated with data obtained from isolated porcine hearts. RESULTS: In healthy tissue, bipolar Vpp values increased with increasing interelectrode spacing and plateaued beyond a spacing of approximately 4 mm. The bipolar Vpp values in healthy tissue were relatively less sensitive to wavefront orientation angle with large interelectrode spacing. In diseased tissue, on the contrary, with increasing interelectrode spacing, bipolar Vpp values increased linearly without a plateau and were more sensitive to orientation angle. The bipolar Vpp values decreased with increasing thickness of the scar, with larger relative decrease in small bipoles than larger ones. Bipolar Vpp values increased with a progressively intramural location of fixed-size scar and became less distinguishable from healthy tissue especially for smaller interelectrode spacings. CONCLUSIONS: The scalable relationship established for interelectrode distances favour an electric-field-based assessment as opposed to traditional Vpp values as a tool for physiologically relevant measurement for mapping catheters with interelectrode spacing up to 4 mm. This will allow for universal assessment of myocardial health across catheters with varied spacing.

Quantifying the determinants of decremental response in critical ventricular tachycardia substrate.

BACKGROUND: Decremental response evoked with extrastimulation (DEEP) is a useful tool for determining diastolic return path of ventricular tachycardia (VT). Though a targeted VT ablation is feasible with this approach, determinants of DEEP response have not been studied OBJECTIVES: To elucidate the effects of clinically relevant factors, specifically, the proximity of the stimulation site to the arrhythmogenic scar, stimulation wave direction, number of channels open in the scar, size of the scar and number of extra stimuli on decrement and entropy of DEEP potentials. METHODS: In a 3-dimensional bi-domain simulation of human ventricular tissue (TNNP cell model), an irregular subendocardial myopathic region was generated. An irregular channel of healthy tissue with five potential entry branches was shaped into the myopathic region. A bipolar electrogram was derived from two electrodes positioned in the centre of the myopathic region. Evoked delays between far-field and local Electrogram (EGM) following an extrastimulus (S1-S2, 500-350 ms) were measured as the stimulation site, channel branches, and inexcitable tissue size were altered. RESULTS: Stimulation adjacent to the inexcitable tissue from the side opposite to the point-of-entry produces longest DEEP delay. The DEEP delay shortens when the stimulation point is farther away from the scar, and it decreases maximally when stimulation is done from a site beside a conduction barrier. Entropy increases with S2 when stimulation site is from farther away. An unprotected channel structure with multiple side-branch openings had shorter DEEP delay compared to a protected channel structure with a paucity of additional side-branch openings and a point-of-entry on the side opposite to the pacing source. Addition of a second shorter extrastimulus did not universally lead to higher DEEP delay CONCLUSIONS: Location and direction of the wavefront in relation to scar entry and size of scar determine the degree of evoked response while the number of extrastimuli has a small additional decremental effect.

Electrophysiologic features of protected channels in late postinfarction patients with and without spontaneous ventricular tachycardia.

PURPOSE: Protected channels of surviving myocytes in late postinfarction ventricular scar predispose to ventricular tachycardia (VT). However, only a few patients develop VT spontaneously. We studied differences in electric remodeling and protected channels in late postinfarction patients with and without spontaneous VT. METHODS: Patients with ischemic cardiomyopathy (ICM) with recurrent sustained monomorphic VT (n = 22) were compared with stable ICM patients without spontaneous VT (control group; n = 5). Left ventricular mapping was performed with a 20-pole catheter. Detailed pace mapping was used to identify channels of protected conduction, and confirmed, when feasible, by entrainment. Anatomical and electrophysiological properties of VT channels and non-VT channels in VT patients and channels in controls were evaluated. RESULTS: Seventy-three (median 3) VTs were inducible in VT patients compared to two (median 0) in controls. The VT channels in VT patients (n = 57, 3 ± 1 per patient) were lengthier (mean ± SEM 53 ± 5 vs. 33 ± 4 vs. 24 ± 8 mm), had longer S-QRS (73 ± 4 vs. 63 ± 3 vs. 44 ± 8 ms), longer conduction time (103 ± 13 vs. 33 ± 4 vs. 24 ± 8 ms), and slower conduction velocity (CV) (0.85 ± 0.21 vs. 1.39 ± 0.20 vs. 1.31 ± 0.41 m/s) than non-VT channels in VT patients (n = 183, 8 ± 6 per patient) (p ≤ 0.01) and channels in controls (n = 46, 9 ± 8 per patient) (p ≤ 0.01). Additionally, non-VT channels in VT patients had longer S-QRS (p = 0.02); however, they were similar in length, conduction time, and CV compared to channels in controls. CONCLUSIONS: Channels supporting VT are lengthier, with longer conduction times and slower CV compared to channels in patients without spontaneous VT. These observations may explain why some ICM patients have spontaneous VT and others do not.

Development of Time- and Voltage-Domain Mapping (V-T-Mapping) to Localize Ventricular Tachycardia Channels During Sinus Rhythm.

BACKGROUND: In ventricular scar, impulse spread is slow because it traverses split and zigzag channels of surviving muscle. We aimed to evaluate scar electrograms to determine their local delay (activation time) and inequality in voltage splitting (entropy), and their relationship to channels. We reasoned that unlike innocuous channels, which are often short with multiple side branches, ventricular tachycardia (VT) supporting channels have very slow impulse spread and possess low entropy because of their longer protected length and relative lack of side-branching. METHODS AND RESULTS: Patients with ischemic cardiomyopathy and multiple VT were studied. In initial mapping stage (16 patients and 58 VTs), left ventricular endocardial mapping was performed in sinus rhythm. Detailed pace mapping was used to identify VT channels and confirmed, when feasible, by entrainment. Scar electrograms were analyzed in time and voltage domains to determine mean activation time, dispersion in activation time, and entropy. Predictive performances of these properties to detect VT channels were tested. In the application stage (7 patients and 20 VTs), these properties were prospectively tested to guide catheter ablation. A mean number of 763±203 sampling points were taken. From 1770 pace maps, 47 channels corresponded to VTs. A combination of scar electrograms with the latest mean activation time and minimum entropy, in a high activation dispersion region, accurately recognized regions containing VT channels (κ=0.89, sensitivity=86%, specificity=100%, positive predictive value=93%, and negative predictive value=100%). Finally, focused ablation within 5-mm rim of the prospective channel regions eliminated 18 of 20 inducible VTs. CONCLUSIONS: Activation time and entropy mapping in the scar accurately identify VT channels during sinus rhythm. The method integrates principles of reentry formation to recognize VT channels without pace mapping or mapping during VT.

High-density mapping of ventricular scar: a comparison of ventricular tachycardia (VT) supporting channels with channels that do not support VT.

BACKGROUND: Surviving myocytes within scar may form channels that support ventricular tachycardia (VT) circuits. There are little data on the properties of channels that comprise VT circuits and those that are non-VT supporting channels. METHODS AND RESULTS: In 22 patients with ischemic cardiomyopathy and VT, high-density mapping was performed with the PentaRay catheter and Ensite NavX system during sinus rhythm. A channel was defined as a series of matching pace-maps with a stimulus (S) to QRS time of ≥40 ms. Sites were determined to be part of a VT channel if there were matching pace-maps to the VT morphology. This was confirmed with entrainment mapping when possible. Of the 238 channels identified, 57 channels corresponded to an inducible VT. Channels that were part of a VT circuit were more commonly located within dense scar than non-VT channels (97% versus 82%; P=0.036). VT supporting channels were of greater length (mean±SEM, 53±5 versus 33±4 mm), had higher longest S-QRS (130±12 versus 82±12 ms), longer conduction time (103±14 versus 43±13 ms), and slower conduction velocity (0.87±0.23 versus 1.39±0.21 m/s) than non-VT channels (P<0.001). Of all the fractionated, late, and very late potentials located in scar, only 21%, 26%, and 29%, respectively, were recorded within VT channels. CONCLUSIONS: High-density mapping shows substantial differences among channels in ventricular scar. Channels supporting VT are more commonly located in dense scar, longer than non-VT channels, and have slower conduction velocity. Only a minority of scar-related potentials participate in the VT supporting channels.

Bipolar electrogram shannon entropy at sites of rotational activation: implications for ablation of atrial fibrillation.

BACKGROUND: The pivot is critical to rotors postulated to maintain atrial fibrillation (AF). We reasoned that wavefronts circling the pivot should broaden the amplitude distribution of bipolar electrograms because of directional information encoded in these signals. We aimed to determine whether Shannon entropy (ShEn), a measure of signal amplitude distribution, could differentiate the pivot from surrounding peripheral regions and thereby assist clinical rotor mapping. METHODS AND RESULTS: Bipolar electrogram recordings were studied in 4 systems: (1) computer simulations of rotors in a 2-dimensional atrial sheet; (2) isolated rat atria recorded with a multi-electrode array (n=12); (3) epicardial plaque recordings of induced AF in hypertensive sheep (n=11); and (4) persistent AF patients (n=10). In the model systems, rotation episodes were identified, and ShEn calculated as an index of amplitude distribution. In humans, ShEn distribution was analyzed at AF termination sites and with respect to complex fractionated electrogram mean. We analyzed rotation episodes in simulations (4 cycles) and animals (rats: 14 rotors, duration 80±81 cycles; sheep: 13 rotors, 4.2±1.5 cycles). The maximum ShEn bipole was consistently colocated with the pivot zone. ShEn was negatively associated with distance from the pivot zone in simulated spiral waves, rats, and sheep. ShEn was modestly inversely associated with complex fractionated electrogram; however, there was no relationship at the sites of highest ShEn. CONCLUSIONS: ShEn is a mechanistically based tool that may assist AF rotor mapping.

Mapping and ablation of the pulmonary veins and cavo-tricuspid isthmus with a magnetic resonance imaging-compatible externally irrigated ablation catheter and integrated electrophysiology system.

BACKGROUND: Magnetic resonance imaging (MRI)-guided interventional electrophysiology (EP) has rapidly emerged as a promising alternative to x-ray-guided ablation. We aimed to evaluate an externally irrigated MRI-compatible ablation catheter and integrated EP pacing and recording system, testing the feasibility of pulmonary vein and cavo-tricuspid isthmus ablation. METHODS AND RESULTS: Externally irrigated MRI-compatible ablation and diagnostic EP catheters and an integrated EP recording system (Imricor Medical Systems, Burnsville, MN) were tested in n=11 sheep in a 1.5-T MRI scanner. Power-controlled (40 W, 120-second duration) lesions were formed at the pulmonary vein and cavo-tricuspid isthmus. Real-time intracardiac electrograms were recorded during MRI. Steady-state free precession non-breath-hold images were repeatedly acquired to guide catheter navigation. Lesion visualization was performed using noncontrast (T2-weighted turbo spin echo pulse sequence) and gadolinium-diethylene triamine pentaacetic acid-enhanced T1-weighted imaging (inversion-recovery gradient echo pulse sequence). Catheters were able to be visualized and navigated under cardiovascular magnetic resonance guidance. In total, 8±2.5 lesions (radiofrequency time, 16±4.2 minutes) were formed at the pulmonary vein ostia, and 6.5±1.3 lesions (radiofrequency time, 13±2.2 minutes) were formed at the cavo-tricuspid isthmus, with the end point of bidirectional block. The mean procedure time was 150±55 minutes. Lesion visualization with both T2W imaging and contrast-enhanced imaging correlated with sites of injury at autopsy. CONCLUSIONS: These data demonstrate the feasibility of using multiple catheters, an integrated EP pacing and recording system, and externally irrigated ablation with cardiovascular magnetic resonance guidance to undertake clinically relevant biatrial mapping and ablation.

Results of radiofrequency ablation of permanent atrial fibrillation of >2 years duration and left atrial size >5 cm using 2-mm irrigated tip ablation catheter and targeting areas of complex fractionated atrial electrograms.

Complex fractionated atrial electrograms (CFAEs) have shown promise as target sites for ablation of atrial fibrillation (AF); however, the data are limited with regard to patients with a large left atrium (LA) (>5 cm), and/or a permanent AF duration of >2 years. We tested the hypothesis that ablation of user-defined, computer-generated CFAE and pulmonary vein isolation, without additional lines would help long-term maintenance of sinus rhythm (SR). A total of 21 patients, 9 men and 12 women, aged 32 to 78 years (mean 44 +/- 3.3) were selected. All had chronic AF for >2 years (range 2 to 20; mean 3.8) and a LA of 5.3 to 11.3 cm (mean 6.4 cm). The underlying structural heart disease was rheumatic mitral valve disease in 18, aortic stenosis in 1, and hypertension in 2. Mapping and ablation was done using the NAVx Ensite system and a 2-mm-tip IBI Therapy Cool Path ablation catheter. The target included circumferential pulmonary vein ablation and elimination of areas in the LA and proximal coronary sinus showing CFAEs. During ablation, 3 patients converted to SR. In 15 others, significant organization of the atrial activity occurred. They then underwent successful electrical cardioversion. Three patients showed no change in atrial activity nor had electrical cardioversion. No procedural complications occurred. Patients took oral amiodarone for 3 months after the procedure. At 3 to 12 months (mean 9.8) of follow-up, 3 patients who were in AF at the end of the ablation procedure continued to be in AF. Of the rest, all but 3 were able to maintain SR without antiarrhythmic drugs. In conclusion, ablation using a 2-mm tip irrigation catheter, targeting user-defined CFAEs and pulmonary vein isolation facilitated maintenance of SR in most patients with a LA >5 cm and an AF duration of >2 years.